Macedo et al. Vet Res (2021) 52:68 https://doi.org/10.1186/s13567-021-00935-9 RESEARCH ARTICLE Open Access Molecular characterization of Glaesserella parasuis strains isolated from North America, Europe and Asia by serotyping PCR and LS‑PCR Nubia Macedo1* , Marcelo Gottschalk2, Katrin Strutzberg‑Minder3, Chao Nguyen Van4,5, Lijun Zhang5, Geng Zou5, Rui Zhou5, Thaire Marostica1,6, Maria Jose Clavijo1,7†, Alexander Tucker8† and Virginia Aragon9,10† Abstract Glaesserella parasuis strains were characterized by serotyping PCR, vtaA virulence marker Leader Sequence (LS)-PCR, clinical significance, and geographic region. Overall, the serovars 4, 5/12, 7, 1, and 13 were the most commonly detected. Serovars of greatest clinical relevance were systemic isolates that had a higher probability of being serovar 5/12, 13, or 7. In comparison, pulmonary isolates had a higher likelihood of being serovars 2, 4, 7, or 14. Serovars 5/12 and 13 have previously been considered disease-associated, but this study agrees with other recent studies showing that serovar 7 is indeed associated with systemic G. parasuis disease. Serovar 4 strains illustrated how isolates can have varying degrees of virulence and be obtained from pulmonary, systemic, or nasal sites. Serovars 8, 9, 15, and 10 were predominantly obtained from nasal samples, which indicates a limited clinical significance of these serovars. Addi‑ tionally, most internal G. parasuis isolates were classified as virulent by LS-PCR and were disease-associated isolates, including serovars 1, 2, 4, 5/12, 7, 13, and 14. Isolates from the nasal cavity, including serovars 6, 9, 10, 11, and 15, were classified as non-virulent by LS-PCR. In conclusion, the distribution of G. parasuis serovars remains constant, with few serovars representing most of the strains isolated from affected pigs. Moreover, it was confirmed that the LS-PCR can be used for G. parasuis virulence prediction of field strains worldwide. Keywords: Glaesserella parasuis, Serotyping, VtaA marker, LS-PCR Introduction fibrinopurulent polyserositis, polyarthritis, and meningi- Glaesserella parasuis is an important pathogen of swine tis, commonly known as Glässer’s disease [1]. and remains one of the most important bacterial causes Glässer’s disease is considered one of the most preva- of mortality in swine production worldwide. Clini- lent swine bacterial infections worldwide [2]. Reliance cal signs of infected pigs range from fever, respiratory on prophylactic or strategic medication to control this signs, and swollen joints with lameness, to central nerv- disease is an ongoing concern regarding the emergence ous signs, and sudden death. Chronic infections are also of antimicrobial resistance [3–6]. Bacterins and a live associated with reduced growth rates and food conver- attenuated vaccine are commercially available. However, sion efficiencies. The main pathological findings include G. parasuis is a highly diverse pathogen, with still poorly known cross-protection against disease-causing field strains [7, 8]. In many countries, autogenous bacterins *Correspondence: nubia@iastate.edu are also produced. † Maria Jose Clavijo, Alexander Tucker and Virginia Aragon contributed G. parasuis is also a commensal of the upper respira- equally to the work 1 tory tract (URT) [9]. Disease outbreaks are associated Veterinary Diagnostic Laboratory, Iowa State University, Ames, IA, USA Full list of author information is available at the end of the article with the introduction of novel G. parasuis strains into an © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​ mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​mmons.​org/​publi​cdoma​in/​ zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Macedo et al. Vet Res (2021) 52:68 Page 2 of 10 immunologically naïve pig population, typically through vaccine candidate selection, improving pig flow manage- the mixing of pigs from different sources. Other predis- ment, and limiting the reliance on prophylactic antimi- posing factors include poor management, high stocking crobials. This study’s objective was to characterize strains density, inadequate ventilation, and coinfections with of G. parasuis from top pork-producing countries in other respiratory pathogens [10, 11]. terms of the relative geographic prevalence of molecular Epidemiological investigation of G. parasuis-associated serovars and their relationship with vtaA markers. disease, and preventive health management, relies on identifying specific G. parasuis serovars colonizing and Materials and methods or causing illness. Fifteen G. parasuis serovars that dif- Isolate collection fer in virulence have been described [12, 13]. However, The 950 G. parasuis isolates in this collection originated healthy pigs can carry serovars commonly isolated from from pigs from different geographical regions; USA systemic sites in the URT, alongside isolates of serovars (n = 223 from 15 different states), Europe (n = 436 from rarely identified in clinical disease [14, 15]. 13 different Member States), China (n = 167 from 12 dif- This broad diversity within the G. parasuis species in ferent provinces), Vietnam (n = 56, from two provinces), the field requires surveillance tools to detect disease-rele- and Canada (n = 68, from Quebec) (Figure 1, Additional vant strains. Whole-genome sequencing enabled detailed file 1). The dates of collection of isolates ranged between analysis of the capsule loci of G. parasuis [16] and the 2014–2018 for USA isolates, 2000–2006, 2015–2016 or development of a multiplex serotyping PCR [17, 18]. This unknown for European isolates, 1999–2007 or unknown PCR reaction uses specific primers based on serovar-spe- for Chinese isolates, 2017 for Vietnamese isolates, and cific variations on the capsule loci, capable of discrimi- 2017–2019 for Canadian isolates (Additional file 1). The nating 14 of the 15 recognized G. parasuis serovars but G. parasuis isolates used in this study came from diag- not serovars 5 and 12. A new G. parasuis serovar-specific nostic and research laboratories that are national leaders PCR scheme containing primers that differentiate sero- in their own countries regarding swine diseases, includ- vars 5 and 12 was subsequently described [19]. However, ing Glässer’s disease. Only isolates with a designated the primer for serovar 12 is based on a hypothetical gene, serovar were included. However, more specific informa- and further investigations are warranted. tion regarding the clinical history and epidemiology was In parallel to molecular serotyping, novel PCR methods not available for most isolates and was not included in were developed to track virulence markers for G. par- this study. asuis isolates. A group of specific virulence-associated Data for US isolates (n = 223) were obtained from the genes, named virulence-associated trimeric autotrans- Iowa State University Veterinary Diagnostic Laboratory porters (vtaA) group 1-translocator, were associated with (ISU VDL), where isolates were confirmed as G. parasuis virulent G. parasuis strains [20]. Additionally, two differ- using Bruker matrix-assisted laser desorption time-of- ent leader sequences were detected in the vtaA genes. A flight mass spectrometry (MALDI_TOF MS). Data for test for predicting the virulence potential of G. parasuis Canadian strains (n = 66) were obtained from the Diag- strains was developed based on the leader sequence of nostic Laboratory of the Faculty of Veterinary Medicine the vtaA genes, called LS-PCR [21]. of the University of Montreal (DLFVMUM), where iso- Such epidemiological investigations are essential lates were also firstly identified by MALDI_TOF MS for designing preventive health programs, supporting and confirmed by PCR [22]. Data for European isolates Figure 1 Distribution of Glaesserella parasuis isolates per country. Macedo et al. Vet Res (2021) 52:68 Page 3 of 10 were obtained from Spain (IRTA-CReSA, n = 275) and [17], but these were tested individually as follows: 94 °C Germany (IVD Innovative Veterinary Diagnostics (IVD for 5 min, followed by 30 cycles of 94 °C for 30 s, 55 °C GmbH), (n = 165) and represented 13 different member for 30 s, and 72 °C for 60 s, and final extension of 72 °C states. Confirmation of G. parasuis by both laborato- for 7 min. PCR products were visualized by agarose gel ries was achieved by PCR [22]. Data for Chinese isolates electrophoresis. (n = 167) and Vietnamese isolates (n = 56) were obtained from Huazhong Agricultural University Veterinary Diag- Virulence gene PCR (LS‑PCR) nostic Laboratory. The Chinese and Vietnamese isolates Isolates from North America and Europe were tested were firstly identified with biochemical tests [23] and using a PCR to detect the leader sequence (LS) of the confirmed by 16S rRNA PCR [25]. vtaA genes domain as a diagnostic tool to predict G. par- Metadata for a subset of these isolates included isola- asuis virulence [21]. Briefly, the PCR reaction consisted tion site, year of isolation, and country of origin (Addi- of 12.5 µL of 2 × Multiplex PCR Master Mix (Qiagen), tional file 1). Metadata for the strains were used to 1 µL of the 100 nmol ­1–1 of each primer, 4.5 µL of nucle- classify isolates considering clinical (internal organs) and ase-free water, and 2 µL of extracted DNA in a final vol- non-clinical (URT) source when such information was ume of 25 μL. Cycling conditions were 15 min at 95 °C, available. Further classification was performed consid- followed by 30 cycles of 45 s at 94 °C, 45 s at 52 °C and ering clinical (internal) isolates that were obtained from 1 min at 72 °C, then a final incubation at 72 °C for 10 min. systemic sites or pulmonary (lungs) and non-clinical (URT) origins. G. parasuis strains isolated from internal Statistical analysis organs other than the lungs, such as pleura, pericardium, Associations between serovars, sample site, country of peritoneum, joint, and meninges, were considered sys- isolation, and LS-PCR results were calculated using Chi- temic strains. All URT strains in this study were isolated square and Fisher’s exact tests. A multinomial logistic from nasal cavities. regression model was fit to compare the sample site and serovar classification to the reference (untypeable or NT) Serotyping PCR serovar. All tests were 2-sided, and a p-value of lower Serotyping PCR was undertaken by each participating than 0.05 was defined as indicating a significant differ- laboratory with some laboratory-specific amendments. ence. All analyses were completed using SAS version 9.4. DNA extraction was achieved by standard methods; typi- cally, a loopful of bacterial colonies were removed from Results growth media and suspended in 300 µL of 0.01 M ster- Distribution of G. parasuis serovars ile phosphate-buffered saline (pH 7.4) without calcium Most of the G. parasuis isolates (886/950, 93.3%), were and magnesium (GIBCO/Life Technologies Corporation, typeable by serotyping PCR, and the remaining 64 extracted with the MagMAX™ Viral DNA/RNA Isolation Carlsbad, CA, USA) and vortexed. Bacterial DNA was (9.7%) were identified as non-typeable (NT). All 15 sero- vars were identified. Serovars 4 (17.4%), 5/12 (16.8%), Kit (Life Technologies, Carlsbad, CA, USA) using extrac- 7 (11.4%), 1 (10.7%), and 13 (9.1%) were the most com- tion protocol per manufacturer instructions. monly detected serovars across the participating regions All participating laboratories conducted the serotyping (Figure 2, Table 1, Additional files 2, 3, 4). PCR as previously described [17] with some (ISU VDL, DLFVMUM, and IVD) using the modifications made by Internal isolates Lacouture et al. [18]. Information on the site of origin for each isolate was Briefly, each PCR reaction (one for each of the primer available for 834 of the total 950 G. parasuis isolates, mixtures) consisted of 12.5 µL of 2 × Multiplex PCR allowing classification of G. parasuis strains as either Master Mix (Qiagen, Hilden, Germany), 2.5 µL of the nasal or internal (Table 2). Internal isolates represented 100 nmol ­1–1 of primer mix 1, 2 or 3, 8 µL of nuclease- 81.1% (676/834) of all isolates for which an attributable free water, and 2 µL of extracted DNA. The reaction was sample site was recorded (Table 2). The most frequently run using the following conditions: first cycle at 95 °C for identified serovars from internal locations included sero- 15 min, followed by 40 cycles of 94 °C for 30 s, 58 °C for vars 4, 5/12, 1, 7, and 13 in descending order, with sero- 90 s, and 72 °C for 90 s. This was followed by one cycle at vars 4 and 5/12 being similarly represented, at 19.2 and 72 °C for 10 min. Electrophoresis and visualization of the 18.2%, respectively. PCR products were performed using QIAxcel Capillary Further classification of internal isolates into pul- Electrophoresis with screen gel software (Qiagen). monary and systemic was possible for a subset of 519 For isolates obtained by IRTA-CReSA, PCR conditions of these isolates. Almost sixty percent of these further were amended using the same primers as Howell et al. classified isolates (307/519, 59.2%) were obtained from Macedo et al. Vet Res (2021) 52:68 Page 4 of 10 180 160 Total # of serotypes idenfied by PCR 140 120 100 80 60 40 20 0 1 2 3 4 5/12 6 7 8 9 10 11 13 14 15 NT Serotypes Figure 2 Frequency distribution of Glaesserella parasuis serovars identified by PCR for all of the isolates in the present study. Table 1 Number and percentage of Glaesserella parasuis serovars per country or region. Serovar USA/% Canada/% Europe/% China/% Vietnam/% Total¶ 1 32/14.3 4/5.9 53/12.2 12/7.2 1/1.8 102 2 28/12.6 4/5.9 13/3 8/4.8 13/23.2 66 3 0/0 1/1.5 0/0 5/3 0/0 6 4 53/23.8 8/11.8 75/17.2 19/11.4 10/17.9 165 5/12 28/12.6 21/30.9 60/13.8 36/21.6 15/26.8 160 6 2/0.9 1/1.5 12/2.8 12/7.2 1/1.8 28 7 41/18.4 11/16.2 51/11.7 4/2.4 1/1.8 108 8 0/0 2/2.9 21/4.8 13/7.8 1/1.8 37 9 0/0 0/0 26/6 7/4.2 6/10.7 39 10 0/0 0/0 18/4.1 6/3.6 7/12.5 31 11 0/0 0/0 3/0.7 1/0.6 0/0 4 13 23/10.3 6/8.8 51/11.7 6/3.6 0/0 86 14 12/5.4 2/2.9 11/2.5 8/4.8 0/0 33 15 1/0.4 1/1.5 18/4.1 1/0.6 0/0 21 NT 3/1.3 7/10.3 24/5.5 29/17.4 1/1.8 64 Total§ 223/100 68/100 436/100 167/100 56/100 950 ¶ Sum of number of serovars in each row § Sum of number of serovars in each column pulmonary sites, while 40.8% (212/519) were of sys- pulmonary and systemic collections (12.1 and 15.6% of temic origin (Table 2). Serovar 4 strains were more fre- each collection, respectively), the percentage of serovar quently isolated from lung samples than serovar 5/12 13 strains was higher in the systemic collection com- strains (27 and 14.7%, respectively). In contrast, when pared with the pulmonary one (14.2 and 9.1%, respec- considering systemic isolates separately, the percentage tively) (Table 2). of serovar 5/12 strains was higher compared with sero- Clinically relevant serovars from internal sites, such var 4 (22.2 and 13.7%, respectively) (Table 2). Likewise, as serovars 4, 5/12, 7, 1, and 13, were among the six both serovars 7 and 13 were more frequently detected serovars more frequently observed in North America within internal samples, than nasal samples. However, and Europe, with similar distribution among systemic while serovar 7 was similarly represented within both and pulmonary sites (Tables 3, 4, 5). Among Asian Macedo et al. Vet Res (2021) 52:68 Page 5 of 10 Table 2 Number and percentage of Glaesserella parasuis while only 1.4% (3/210) of USA internal isolates were serovars per site of isolation. untypable (Table 3). Serovar Nasal/% Internal/% Pulmonary*/% Systemic*/% Total¶ 1 19/12 75/11.1 37/12.1 29/13.7 94 Nasal isolates 2 6/3.8 53/7.8 31/10.1 14/6.6 59 Nasal isolates represented 18.1% (158/834) of all iso- 3 0/0 6/0.9 0/0 1/0.5 6 lates for which an attributable sample site was recorded 4 18/11.4 130/19.2 83/27 29/13.7 148 and belonged to a variety of serovars, especially sero- 5/12 13/8.2 124/18.3 45/14.7 47/22.2 137 vars 8 (12%), 1 (12%), 4 (11.4%), 9 (11.4%), 15 (10.8%) 6 7/4.4 19/2.8 2/0.7 5/2.4 26 and 10 (9.5%) (Table 2). While serovars 1 and 4 were 7 8/5.1 74/10.9 37/12.1 33/15.6 82 more consistently found within systemic or pulmo- 8 19/12 16/2.4 3/1 1/0.5 35 nary sites, serovars 8, 9, 15, and 10 were predominantly 9 18/11.4 17/2.5 6/2 4/1.9 35 obtained from nasal samples (Table 2). 10 15/9.5 14/2.1 7/2.3 1/0.5 29 Most of the nasal isolates in this study were obtained 11 0/0 4/0.6 2/0.7 1/0.5 4 from Europe (76.6%, 121/158) (Table 6). The most prev- 13 8/5.1 63/9.3 28/9.1 30/14.2 71 alent nasal serovars in Europe were 8, 15, 1, 4, 9, and 10. 14 1/0.6 30/4.4 15/4.9 7/3.3 31 Most of these serovars were not prevalent among inter- 15 17/10.8 2/0.3 0/0 1/0.5 19 nal samples, except for serovars 1 and 4, which were NT 9/5.7 49/7.2 11/3.6 9/4.2 58 also prevalent among pulmonary and systemic sites Total§ 158/100 676/100 307/100 212/100 834 (Tables 3, 4, 5, 6). *Information was available for 519 internal isolates, which were divided into Besides these European nasal isolates, 35 G. paras- pulmonary and systemic uis nasal isolates were from Asian countries, includ- ¶ Sum of nasal and internal isolates in each row ing 25 from Vietnam and ten from China. Most of § Sum of values in each column the nasal isolates from China were either from sero- var 5/12 or 1, while most of the nasal isolates from Vietnam were from serovars 2, 9, 4, and 10 (Tables 6). isolates, serovar 5/12 was the most prevalent, followed However, only serovars 1 and 9 were predominantly by serovars 4 and 2 (Table 3). higher among nasal isolates. The other serovars The highest proportion of untypable internal isolates were also frequently found among internal isolates (18.5%, 29/157) was found among Chinese isolates, (Table 3). Table 3 Number and percentage of Internal Glaesserella parasuis serovars per country or region. Serovar USA/% Canada/% Europe/% China/% Vietnam/% Total¶ 1 31/14.8 3/6.7 32/13.7 9/5.7 0/0 75 2 27/12.9 3/6.7 8/3.4 8/5.1 7/22.6 53 3 0/0 1/2.2 0/0 5/3.2 0/0 6 4 51/24.3 1/2.2 54/23.2 18/11.5 6/19.4 130 5/12 28/13.3 15/33.3 35/15 32/20.4 14/45.2 124 6 2/1 1/2.2 4/1.7 12/7.6 0/0 19 7 34/16.2 8/17.8 28/12 4/2.5 0/0 74 8 0/ 1/2.2 3/1.3 12/7.6 0/0 16 9 0/0 0/0 10/4.3 7/4.5 0/0 17 10 0/0 0/0 4/1.7 6/3.8 4/42.9 14 11 0/0 0/0 3/1.3 1/0.6 0/0 4 13 22/10.5 4/8.9 32/13.7 5/3.2 0/0 63 14 12/5.7 1/2.2 9/3.9 8/5.1 0/0 30 15 0/0 1/2.2 0/0 1/0.6 0/0 2 NT 3/1.4 6/13.3 11/4.7 29/18.5 0/0 49 Total§ 210/100 45/100 233/100 157/100 31/100 676 ¶ Sum of number of serovars in each row. § Sum of number of serovars in each column. Macedo et al. Vet Res (2021) 52:68 Page 6 of 10 Table 4 Number and percentage of Pulmonary Glaesserella Table 6 Number and percentage of Nasal Glaesserella parasuis serovars per country or region. parasuis serovars per country or region. Serovar USA/% Europe/% Vietnam/% Total¶ Serovar USA/% Europe/% China/% Vietnam/% Total¶ 1 15/11.9 22/14.7 0/0 37 1 1/50 14/11.6 3/30 1/4 19 2 17/13.5 7/4.7 7/22.6 31 2 0/0 0/0 0/0 6/24 6 4 39/31 38/25.3 6/19.4 83 4 1/50 12/9.9 1/10 4/16 18 5/12 15/11.9 16/10.7 14/45.2 45 5/12 0/0 8/6.6 4/40 1/4 13 6 1/0.8 1/0.7 0/0 2 6 0/0 6/5 0/0 1/4 7 7 19/15.1 18/12 0/0 37 7 0/0 7/5.8 0/0 1/4 8 8 0/0 3/2 0/0 3 8 0/0 17/14 1/10 1/4 19 9 0/0 6/4 0/0 6 9 0/0 12/9.9 0/0 6/24 18 10 0/0 3/2 4/12.9 7 10 0/0 12/9.9 0/0 3/12 15 11 0/0 2/1.3 0/0 2 13 0/0 7/5.8 1/10 0/0 8 13 7/5.6 21/14 0/0 28 14 0/0 1/0.8 0/0 0/0 1 14 11/8.7 4/2.7 0/0 15 15 0/0 17/14 0/0 0/0 17 NT 2/1.6 9/6 0/0 11 NT 0/0 8/6.6 0/0 1/4 9 Total§ 126/100 150/100 31/100 307 Total§ 2/100 121/100 10/100 25/100 158 ¶ ¶ Sum of number of serovars in each row. Sum of number of serovars in each row. § § Sum of number of serovars in each column. Sum of number of serovars in each column. There is no information whether Canadian and Chinese internal isolates There were no nasal G. parasuis isolates submitted from Canada. originated from pulmonary sites. Association between Glaesserella parasuis serovars and site Table 5 Number and percentage of Systemic Glaesserella of isolation parasuis serovars per country or region. Results of the chi-square analysis showed an associa- Serovar USA/% Canada/% Europe/% Total ¶ tion between sample type (nasal/internal) and serovar (p < 0.001). A multinomial logistic regression model was 1 16/19 3/6.7 10/12 29 further used to predict serovar classification based on 2 10/11.9 3/6.7 1/1.2 14 sample type. Given the sample type (nasal/internal), sam- 3 0/0 1/2.2 0/0 1 ples collected from the nasal cavity had a higher prob- 4 12/14.3 1/2.2 16/19.3 29 ability of being of serovars 8, 9, 10, or 15 (p < 0.05), but 5/12 13/15.5 15/33.3 19/22.9 47 no association was found between serovar and internal 6 1/1.2 1/2.2 3/3.6 5 samples, during this initial analysis. 7 15/17.9 8/17.8 10/12 33 Using the same model, internal isolates were classified 8 0/0 1/2.2 0/0 1 as either pulmonary or systemic and compared against 9 0/0 0/0 4/4.8 4 nasal isolates for the subsequent analysis. Given the 10 0/0 0/0 1/1.2 1 sample type, systemic isolates had a higher probability 11 0/0 0/0 1/1.2 1 of being of serovar 5/12, 13, or 7 compared to the nasal 13 15/17.9 4/8.9 11/13.3 30 sample collection (p < 0.01). In comparison, pulmonary 14 1/1.2 1/2.2 5/6 7 isolates had a higher likelihood of being serovars 2, 4, 7, 15 0/0 1/2.2 0/0 1 or 14 (p < 0.01). No statistical significance was observed NT 1/1.2 6/13.3 2/2.4 9 when comparing systemic and pulmonary isolates. These Total§ 84/100 45/100 83/100 212 results agree with Olvera et al. [20], where pulmonary ¶ Sum of number of serovars in each row. isolates could be either virulent or non-virulent since § Sum of number of serovars in each column. the lung can be populated by upper respiratory tract There is no information whether Asian internal isolates originated from systemic commensal microbiota and strains that harbor certain sites. virulence factors that enable them to survive the lung defenses and invade systemically. As shown in Table 2, 14/53 and 7/31 of serovars 2 and 14, respectively, were also found on systemic sites. Macedo et al. Vet Res (2021) 52:68 Page 7 of 10 Association between Glaesserella parasuis serovars Table 8 Distribution of Glaesserella parasuis serovars by site and LS‑PCR result of isolation and LS-PCR results. G. parasuis isolates were classified as potentially viru- LS-PCR Systemic Pulmonary Nasal P-value lent based on the presence or absence of specific leader sequences within their vtaA genes, using multiplex LS- Serovar 1 Virulent 29 37 15 PCR. Within this G. parasuis collection, 727 isolates were Non-Viru- 0 0 0 Not Tested* lent tested by LS-PCR among North American and Euro- Serovar 2 Virulent 14 23 0 pean isolates (Table 7). Overall, 85.7% (623/727) of the G. Non-Viru- 0 1 0 0.439 parasuis isolates were classified as virulent by LS-PCR, lent 14.3% (104/727) were classified as non-virulent by LS- Serovar 3 Virulent 0 0 0 PCR, and 223 isolates were not tested (Table 7). Non-Viru- 1 0 0 Not Tested Most of the disease-associated isolates, including 1, 2, lent 4, 5/12, 7, 13, and 14, were considered virulent by LS- Serovar 4 Virulent 14 23 0 PCR, with few exceptions where these virulent serovars Non-Viru- 0 1 0 0.439 were deemed to be non-virulent by LS-PCR (10/578) lent (Table 7). In contrast, most of the isolates from the nasal Serovar Virulent 47 30 6 5/12 cavity, including serovars 6, 9, 10, and 15, were consid- Non-Viru- 0 1 2 0.002 ered non-virulent by LS-PCR (Table 8), with a significant lent association between serovar and classification as virulent Serovar 6 Virulent 1 0 2 or non-virulent by LS-PCR (p < 0.001). Further analy- Non-Viru- 4 2 4 0.612 sis was performed within each serovar to evaluate the lent association of virulent and non-virulent strains with the Serovar 7 Virulent 33 37 5 site of isolation. A significant association was found for Non-Viru- 0 0 2 0.000 serovars 5/12 and 7, which had a higher chance of being lent virulent and from internal sites (systemic or pulmonary). Serovar 8 Virulent 0 0 0 Contrarily, serovar 10 had a higher chance of being non- Non-Viru- 1 3 17 Not Tested virulent and from a nasal site (Table 8). lent Serovar 9 Virulent 0 3 1 Non-Viru- 4 3 11 0.056 lent Serovar 10 Virulent 1 1 0 Table 7 Number and percentage of virulent and non- Non-Viru- 0 2 12 0.007 virulent isolates within the same serovar. lent Serovar Virulent by Non-Virulent Not Tested by Total¶ Serovar 11 Virulent 1 2 0 LS-PCR/% by LS-PCR/% LS-PCR/% Non-Viru- 0 0 0 Not Tested lent 1 89/87.3 0/0 13/12.7 102 Serovar 13 Virulent 30 26 6 2 44/66.7 1/1.5 21/31.8 66 Non-Viru- 0 2 1 0.188 3 0/0 1/16.7 5/83.3 6 lent 4 136/82.4 0/0 29/17.6 165 Serovar 14 Virulent 7 15 1 5/12 106/66.3 3/1.9 51/31.9 160 Non-Viru- 0 0 0 Not Tested 6 4/14.3 11/39.3 13/46.4 28 lent 7 101/93.5 2/1.9 5/4.6 108 Serovar 15 Virulent 0 0 2 8 0/0 23/62.2 14/37.8 37 Non-Viru- 1 0 15 0.716 lent 9 6/15.4 20/51.3 13/33.3 39 *Not Tested: All isolates from this serovar had the same LS-PCR result. 10 2/6.5 16/51.6 13/41.9 31 P-value < 0.05 was considered significant. 11 3/75 0/0 1/25 4 13 77/89.5 3/3.5 6/7 86 14 25/75.8 0/0 8/24.2 33 Discussion 15 3/14.3 17/81 1/4.8 21 Serotyping has been a necessary means of character- NT 27/42.2 7/10.9 30/46.9 64 izing G. parasuis isolates. Using a novel serotyping Total§ 623/65.6 104/10.9 223/23.5 950 PCR method [17, 18], we described the distribution of ¶ Sum of number of serovars in each row. G. parasuis serovars globally and within a geographic § Sum of number of serovars in each column. region. The data collected in this study are from different Macedo et al. Vet Res (2021) 52:68 Page 8 of 10 diagnostic laboratories and research centers. Even challenge studies as indicators of virulence. Isolates of though they serve as centers of expertise on the diagnosis serovar 10 have been reported as disease and non-disease of swine diseases, the dataset used in this study is lim- associated [31]. Still, while the present study found sero- ited to isolates available in these laboratories and, there- var 10 in 9.9% of nasal samples from Europe (the region fore, do not represent a true prevalence of G. parasuis supplying the most significant bulk of nasal samples), it serovars. Additionally, some of the strains from Europe was found in only 1.7% of internal samples from Europe – (57/436) and China (167/223) were isolated more than along with a small cluster of internal isolates from China ten years ago or have an unknown date of isolation. Thus, and Vietnam. Thus, the present study indicates only lim- the actual distribution (prevalence) of G. parasuis sero- ited significance for this serovar. vars is still unknown. Even though our data do not rep- Nevertheless, pigs harbor various G. parasuis strains resent prevalence, most of these isolates originate from in their URT, and potentially pathogenic serovars can field cases and reflect strains frequently found in disease- be isolated from the URT of healthy pigs [9, 14]. In this associated cases, and thus useful information for swine study, URT isolates originated mainly from Europe and practitioners and researchers. Asia, representing 16.6% of this study’s collection. Sero- The serovars’ overall distribution varied slightly among vars 6, 8, 9, 10, and 15 were amongst the most commonly countries, and four serovars (1, 4, 5/12, and 7) accounted identified (43.6%) in the URT. In comparison, these for half of all isolates. Serovars 4 and 5/12, in particular, serovars represented only 7.2% of all internal isolates. were consistently represented across all regions in this Serovars 10 and 15 have been traditionally considered study and previous studies [13, 22–28]. virulent, even though they are not prevalent as disease- The serovars of utmost clinical significance were identi- associated serovars worldwide, and were classified as fied in samples from internal sites (serovars 4, 5/12, 1, 7, non-virulent by the LS-PCR. Interestingly, serovars 6 and and 13) in this study. These serovars have been previously 8 Asian isolates were also frequently isolated from Chi- considered disease-associated [12, 13, 22–30]. Variable nese samples’ internal sites, representing an increase of sample sizes limited our ability to discern significant dif- more virulent strains belonging to these serovars in Asia. ferences in serovar prevalence at the country level. The Different pathogenic properties have also been relative scarcity of serovar 4 isolates among Canada’s described for G. parasuis serovar 4 strains. The refer- internal samples was notable, and serovars 1, 7, and 13 ence isolate for serovar 4 was initially isolated from the appeared relatively similarly scarce among internal sam- nose of a healthy pig and considered mildly virulent ples from China and Vietnam. [12]. In our study, serovar 4 isolates were mainly associ- An important finding in this present study and other ated with pulmonary sites (83/130, 64%), followed by recent studies [28–30] was that serovar 7 is a significant systemic (29/130) and nasal sites (18/130). In an earlier disease-associated serovar of G. parasuis. In this study, study, North American isolates of serovar 4 were mainly serovar 7 was assigned to 10.9% of all internal samples isolated from lung samples of pigs with pneumonia and (ranked ­4th most prevalent serovar) and 15.6% of all other systemic sites, but not from URT samples [22]. systemic isolates (ranked second most prevalent sero- When two serovar 4 European field isolates were used to var). Serovar 7 was initially found non-virulent by intra- infect pigs experimentally, one could reproduce Glasser’s peritoneal challenge [12], and it was the most prevalent disease, while the other could not [31]. These results are serovar identified from the URT of healthy pigs in China consistent with other reports showing serovar 4 isolates [14]. Conversely, a serovar 7 field strain was able to cause clustering with virulent and non-virulent genotypes [31– disease in one snatch-farrowed colostrum-deprived pig 33]. These findings illustrate how isolates of the same when inoculated intranasally and was consistently iso- serovar can have different degrees of virulence. Addition- lated from pulmonary and systemic locations from pigs ally, some G. parasuis serovars’ virulence may have been affected by Glasser’s disease in Germany [28]. Addition- wrongly assigned by the Kielstein and Rapp-Gabrielson ally, the serovar 7 reference strain 174 caused disease in scheme [12]. colostrum-deprived and commercial pigs after experi- Glasser’s disease is a complex multifactorial disease, mental inoculation [29, 30], demonstrating this serovar’s and several different putative virulence genes have been pathogenic potential [31]. identified [34–37]. Still, little experimental data is avail- Based on their detection in fewer than 1% of all inter- able to correlate these virulence genes with clinical dis- nal samples, the least clinical significance serovars were ease. However, empirical evidence exists, demonstrating serovars 3, 11, and 15. Serovars 3 and 11 were previously the group 1 vtaA genes’ role in G. parasuis virulence [21, considered non-virulent [12]. Still, the apparent limited 38, 39]. Therefore, to help identify virulent G. parasuis clinical significance of serovar 15 contradicted previous findings [12], highlighting the limitations of small-scale Macedo et al. Vet Res (2021) 52:68 Page 9 of 10 strains, a new PCR, based on the leader sequence of the Supplementary Information vtaA genes, was developed [21]. The online version contains supplementary material available at https://​doi.​ In the present study, most internal isolates were iden- org/​10.​1186/​s13567-​021-​00935-9. tified as virulent by LS-PCR, while most nasal isolates were non-virulent. Our results agree with previous Additional file 1. Metadata of Glaesserella parasuis isolates including reports [25, 28, 39–41] and supported our disease asso- isolation site, year of isolation, and country of origin. ciation observations for serovar 7 isolates. In our study, Additional file 2. Distribution of Glaesserella parasuis serovars in most serovar 7 (75/77, 97.4%) strains were internal and North America. classified as virulent by the LS-PCR. Only two serovar Additional file 3. Distribution of Glaesserella parasuis serovars in 7 LS-PCR negatives were found in the collection, and Europe. these were from nasal samples. In vitro analysis was Additional file 4. Distribution of Glaesserella parasuis serovars in Asia. not performed in this study to further characterize the virulence of the strains. Previous genome analysis per- formed on a serovar 7 reference strain classified as vir- Acknowledgements We would like to acknowledge the Iowa State University Diagnostic Labora‑ ulent by LS-PCR, but susceptible to both phagocytosis tory for providing isolates and assistance with testing development. We thank and serum demonstrated a reduction in the number of Dr Xuwang Cai (Huazhong Agricultural University) for providing the Chinese vtaAs in that strain, which may represent a strain with clinical isolates, and Nuria Galofre-Mila and Catalina Cruañas-Paniker for their assistance with testing at the IRTA-CReSA. an intermediate phenotype [21]. Nevertheless, some discordant results were also Authors’ contributions found within the LS-PCR data. We found that a small All authors participated in collecting the data, performing the tests, and ana‑ lyzing the data. NM, MJC, MG, KSM, AT, and VA designed the study, analyzed number of isolates from typically non-disease-asso- the data and drafted the manuscript. All authors read and approved the final ciated serovars (such as 6, 8, and 9), and classified as manuscript. non-virulent by LS-PCR, were recovered from systemic Availability of data and materials sites. In this case, other environmental factors or coin- All data generated or analyzed during this study are included in this published fections could be responsible for creating the necessary article and its supplementary information files. conditions for less virulent strains to invade systemi- cally [21, 31, 40]. Even though vtaA genes are positively Declarations correlated with G. parasuis virulent strains and can be Competing interests used as a predictor of virulence, other factors may also The authors declare that they have no competing interests. be necessary for disease development [25, 32, 40]. In conclusion, the overall distribution of G. parasuis Author details 1 Veterinary Diagnostic Laboratory, Iowa State University, Ames, IA, USA. serovars remains predominantly constant over the past 2 Faculty of Veterinary Medicine, University of Montreal, Montreal, Canada. 20 years, with few serovars representing most of the 3 Innovative Veterinary Diagnostics (IVD GmbH), Albert‑Einstein‑Str. 5, strains isolated from Glasser’s disease-affected pigs and 30926 Seelze, Germany. 4 Faculty of Animal Science and Veterinary Medicine, University of Agricultural and Forestry, Hue University, Hue 53000, Vietnam. noting the confirmation of serovar 7 as a globally rel- 5 State Key Laboratory of Agricultural Microbiology, Huazhong Agricultural evant serovar. Even though the serovar and virulence University College of Veterinary Medicine, Wuhan 430070, China. 6 Department relationship is not definitive, serotyping is still suitable of Veterinary Clinic and Surgery, Federal University of Minas Gerais, Av. Antônio Carlos, 6627, Belo Horizonte, MG, Brazil. 7 PIC North America, Hendersonville, for the epidemiological characterization of G. parasuis TN, USA. 8 Department of Veterinary Medicine, University of Cambridge, in pig farms, helping implement more adequate con- Madingley Road, Cambridge CB3 OES, UK. 9 IRTA, Centre de Recerca en Sanitat trol measures, such as selecting isolates to be included Animal (CReSA, IRTA-UAB), Campus de la Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. 10 OIE Collaborating Centre for the Research and Con‑ in vaccines. When serotyping is not available, the site trol of Emerging and Re-Emerging Swine Diseases in Europe (IRTA-CReSA), of isolation can become crucial for screening disease- Bellaterra, Barcelona, Spain. associated isolates since systemic strains are likely the Received: 24 December 2020 Accepted: 13 April 2021 ones that possess the necessary virulent mechanisms for disease development. To date, vtaA genes are the most well-characterized G. parasuis virulence fac- tors. This survey of global G. parasuis isolates confirms References that the LS-PCR is useful for G. parasuis virulence 1. Aragon V, Segales J, Tucker AW (2019) Glässer’s disease. In: Zimmerman prediction. JJ, Karriker LA, Ramirez A, Schwartz K, Stevenson GW, Zhang J (eds) Diseases of Swine. Ames, Wiley-Blackwell Macedo et al. Vet Res (2021) 52:68 Page 10 of 10 2. Zhang B, Tang C, Liao M, Yue H (2014) Update on the pathogenesis of 22. Oliveira S, Blackall PJ, Pijoan C (2003) Characterization of the diversity of Haemophilus parasuis infection and virulence factors. Vet Microbiol 168:1–7 Haemophilus parasuis field isolates by use of serotyping and genotyping. 3. Martín de la Fuente AM, Tucker AW, Navas J, Blanco M, Morris SJ, Gutiérrez- Am J Vet Res 64:435–442 Martín CB (2007) Antimicrobial susceptibility patterns of Haemophilus 23. Cai X, Chen H, Blackall PJ, Yin Z, Wang L, Liu Z, Jin M (2005) Serological parasuis from pigs in the United Kingdom and Spain. Vet Microbiol characterization of Haemophilus parasuis isolates from China. Vet Microbiol 120:184–191 111:231–236 4. Moleres J, Santos-López A, Lázaro I, Labairu J, Prat C, Ardanuy C, González- 24. Rúbies X, Kielstein P, Costa L, Riera P, Artigas C, Espuña E (1999) Prevalence Zorn B, Aragon V, Garmendia J (2015) Novel blaROB-1-bearing plasmid of Haemophilus parasuis serovars isolated in Spain from 1993 to 1997. Vet conferring resistance to β-lactams in Haemophilus parasuis isolates from Microbiol 66:245–248 healthy weaning pigs. Appl Environ Microbiol 81:3255–3267 25. Van CN, Thanh TVT, Zou G, Jia M, Wang Q, Zhang L, Ding W, Huang Q, Zhou 5. Brogden S, Pavlović A, Tegeler R, Kaspar H, De Vaan N, Kehrenberg C (2018) R (2019) Characterization of serotypes and virulence genes of Haemophilus Antimicrobial susceptibility of Haemophilus parasuis isolates from Germany parasuis isolates from Central Vietnam. Vet Microbiol 230:117–122 by use of a proposed standard method for harmonized testing. Vet Micro‑ 26. Dijkman R, Wellenberg GJ, van der Heijden HM, Peerboom R, Olvera biol 217:32–35 A, Rothkamp A, Peperkamp K, van Esch EJ (2012) Analyses of Dutch 6. Zhang B, Ku X, Yu X, Sun Q, Wu H, Chen F, Zhang X, Guo L, Tang X, He Q Haemophilus parasuis isolates by serotyping, genotyping by ERIC-PCR and (2019) Prevalence and antimicrobial susceptibilities of bacterial pathogens Hsp60 sequences and the presence of the virulence associated trimeric in Chinese pig farms from 2013 to 2017. Sci Rep 9:9908 autotransporters marker. Res Vet Sci 93:589–595 7. Miniats OP, Smart NL, Rosendal S (1991) Cross protection among Hae- 27. Ma L, Wang L, Chu Y, Li X, Cui Y, Chen S, Zhou J, Li C, Lu Z, Liu J, Liu Y (2016) mophilus parasuis strains in immunized gnotobiotic pigs. Can J Vet Res Characterization of Chinese Haemophilus parasuis isolates by traditional 55:37–41 serotyping and molecular serotyping methods. PLoS One 11:e0168903 8. Takahashi K, Naga S, Yagihashi T, Ikehata T, Nakano Y, Senna K, Maruyama 28. Schuwerk L, Hoeltig D, Waldmann KH, Strutzberg-Minder K, Valentin-Wei‑ T, Murofushi J (2001) A cross-protection experiment in pigs vaccinated gand P, Rohde J (2020) Serotyping and pathotyping of Glaesserella parasuis with Haemophilus parasuis serovars 2 and 5 bacterins, and evaluation of isolated 2012–2019 in Germany comparing different PCR-based methods. a bivalent vaccine under laboratory and field conditions. J Vet Med Sci Vet Res 51:137 63:487–491 29. Dazzi C, Guizzo J, Prigol S, Kreutz L, Driemeier D, Chaudhuri S, Schryvers A, 9. Cerdà-Cuéllar M, Naranjo JF, Verge A, Nofrarías M, Cortey M, Olvera A, Frandoloso R (2020) New pathological lesions developed in pigs by a “Non- Segalés J, Aragon V (2010) Sow vaccination modulates the colonization of virulent” strain of Glaesserella parasuis. Front Vet Sci 7:98 piglets by Haemophilus parasuis. Vet Microbiol 145:315–320 30. Frandoloso R, Chaudhuri S, Frandoloso G, Yu R, Schryvers A (2020) Proof of 10. Palzer A, Ritzmann M, Wolf G, Heinritzi K (2008) Associations between concept for prevention of natural colonization by oral needle-free adminis‑ pathogens in healthy pigs and pigs with pneumonia. Vet Rec. https://​doi.​ tration of a microparticle vaccine. Front Immunol 11:595320 org/​10.​1136/​vr.​162.9.​267 31. Aragon V, Cerdà-Cuéllar M, Fraile L, Mombarg M, Nofrarías M, Olvera A, 11. Li J, Wang S, Li C, Wang C, Liu Y, Wang G, He X, Hu L, Liu Y, Cui M, Bi C, Shao Sibila M, Solanes D, Segalés J (2010) Correlation between clinico-pathologi‑ Z, Wang X, Xiong T, Cai X, Huang L, Weng C (2017) Secondary Haemophilus cal outcome and typing of Haemophilus parasuis field strains. Vet Microbiol parasuis infection enhances highly pathogenic porcine reproductive and 142:387–393 respiratory syndrome virus (HP-PRRSV) infection-mediated inflammatory 32. Boerlin P, Poljak Z, Gallant J, Chalmers G, Nicholson V, Soltes GA, MacInnes JI responses. Vet Microbiol 204:35–42 (2013) Genetic diversity of Haemophilus parasuis from sick and healthy pigs. 12. Kielstein P, Rapp-Gabrielson VJ (1992) Designation of 15 serovars of Haemo- Vet Microbiol 167:459–467 philus parasuis on the basis of immunodiffusion using heat-stable antigen 33. Wang L, Ma L, Liu Y, Gao P, Li Y, Li X, Liu Y (2016) Multilocus sequence typ‑ extracts. J Clin Microbiol 30:862–865 ing and virulence analysis of Haemophilus parasuis strains isolated in five 13. Angen O, Svensmark B, Mittal KR (2004) Serological characterization of Dan‑ provinces of China. Infect Genet Evol 44:228–233 ish Haemophilus parasuis isolates. Vet Microbiol 103:255–258 34. Sack M, Baltes N (2009) Identification of novel potential virulence-associ‑ 14. Zhang P, Zhang C, Aragon V, Zhou X, Zou M, Wu C, Shen Z (2019) Investiga‑ ated factors in Haemophilus parasuis. Vet Microbiol 136:382–386 tion of Haemophilus parasuis from healthy pigs in China. Vet Microbiol 35. Zhang J, Xu C, Guo L, Shen H, Deng X, Ke C, Ke B, Zhang B, Li A, Ren T, Liao 231:40–44 M (2012) Prevalence and characterization of genotypic diversity of Haemo- 15. Mahmmod YS, Correa-Fiz F, Aragon V (2020) Variations in association of philus parasuis isolates from southern China. Can J Vet Res 76:224–229 nasal microbiota with virulent and non-virulent strains of Glaesserella 36. Zhang NZ, Chu YF, Gao PC, Zhao P, He Y, Lu ZX (2012) Immunological iden‑ (Haemophilus) parasuis in weaning piglets. Vet Res 51:7 tification and characterization of extracellular serine protease-like protein 16. Howell KJ, Weinert LA, Luan SL, Peters SE, Chaudhuri RR, Harris D, Angen encoded in a putative espP2 gene of Haemophilus parasuis. J Vet Med Sci Ø, Aragon V, Parkhill J, Langford PR, Rycroft AN (2013) Gene content and 74:983–987 diversity of the loci encoding biosynthesis of capsular polysaccharides 37. Martínez-Moliner V, Soler-Llorens P, Moleres J, Garmendia J, Aragon V (2012) of the 15 serovar reference strains of Haemophilus parasuis. J Bacteriol Distribution of genes involved in sialic acid utilization in strains of Haemo- 195:4264–4273 philus parasuis. Microbiology 158:2117–2124 17. Howell KJ, Peters SE, Wang J, Hernandez-Garcia J, Weinert LA, Luan SL, 38. Olvera A, Pina S, Pérez-Simó M, Oliveira S, Bensaid A (2010) Virulence- Chaudhuri RR, Angen Ø, Aragon V, Williamson SM, Parkhill J, Langford PR, associated trimeric autotransporters of Haemophilus parasuis are antigenic Rycroft AN, Wren BW, Maskell DJ, Tucker AW (2015) Development of a mul‑ proteins expressed in vivo. Vet Res 41:26 tiplex PCR assay for rapid molecular serotyping of Haemophilus parasuis. J 39. Costa-Hurtado M, Ballester M, Galofré-Milà N, Darji A, Aragon V (2012) Clin Microbiol 53:3812–3821 VtaA8 and VtaA9 from Haemophilus parasuis delay phagocytosis by alveolar 18. Lacouture S, Rodriguez E, Strutzberg-Minder K, Gottschalk M (2017) Sero‑ macrophages. Vet Res 43:57 typing of Haemophilus parasuis field isolates from diseased pigs in Quebec 40. Turni C, Singh R, Blackall PJ (2018) Virulence-associated gene profiling, DNA by indirect hemagglutination assay and multiplex polymerase chain reac‑ fingerprinting and multilocus sequence typing of Haemophilus parasuis tion (PCR). Can Vet J 58:802–804 isolates in Australia. Aust Vet J 96:196–202 19. Jia A, Zhou R, Fan H, Yang K, Zhang J, Xu Y, Wang G, Liao M (2017) Develop‑ 41. Lin WH, Shih HC, Lin CF, Yang CY, Chang YF, Lin CN, Chiou MT (2018) ment of serotype-specific PCR assays for typing of Haemophilus parasuis Molecular serotyping of Haemophilus parasuis isolated from diseased pigs isolates circulating in southern China. J Clin Microbiol 55:3249–3257 and the relationship between serovars and pathological patterns in Taiwan. 20. Olvera A, Pina S, Macedo N, Oliveira S, Aragon V, Bensaid A (2012) Identifica‑ PeerJ 6:e6017 tion of potentially virulent strains of Haemophilus parasuis using a multiplex PCR for virulence-associated autotransporters (vtaA). Vet J 191:213–218 21. Galofré-Milà N, Correa-Fiz F, Lacouture S, Gottschalk M, Strutzberg-Minder Publisher’s Note K, Bensaid A, Pina-Pedrero S, Aragon V (2017) A robust PCR for the differ‑ Springer Nature remains neutral with regard to jurisdictional claims in published entiation of potential virulent strains of Haemophilus parasuis. BMC Vet Res maps and institutional affiliations. 13:124